Discussion

1. Streaming statistics and validation (M1 from Feature Validation RFC)
2. Batch statistics and validation (M2 from Feature Validation RFC)
3. Add support for metadata about missing feature values (#278, Missing Features Metadata RFC)
4. User authentication & authorization (#504)
5. Add feature or feature set descriptions (#463)
6. Redis Cluster Support (#478)

1. Remove feature set versions from API for retrieval only (#462)
2. Tracking of batch ingestion by with dataset_id/job_id (#461)

1. Extended discovery API/SDK (needs to be scoped
   1. Resource listing
   2. Schemas, statistics, metrics
   3. Entities as a higher-level concept (#405)
   4. Add support for discovery based on annotations/labels/tags for easier filtering and discovery
2. Add support for default values (needs to be scoped)
3. Add support for audit logs (needs to be scoped)
4. Support for an open source warehouse store or connector (needs to be scoped)

1. Move all non-registry functionality out of Feast Core and make it optional (needs to be scoped)
   1. Allow Feast serving to use its own local feature sets (files)
   2. Move job management to Feast serving
   3. Move stream management (topic generation) out of Feast core
2. Remove feature set versions from Feast (not just retrieval API) (needs to be scoped)
   1. Allow for auto-migration of data in Feast
   2. Implement interface for adding a managed data store
3. Multi-store support for serving (batch and online) (needs to be scoped)